1. Technical Field
The present invention relates in general to distributed data processing systems and in particular to mechanisms for responding to a communication loss within a virtual input/output server (VIOS) cluster of the data processing system. Still more particularly, the present invention relates to an improved method and system for autonomously migrating a live partition to a new node of the VIOS cluster due to a communication loss within the VIOS cluster.
2. Description of the Related Art
Large scale, distributed data processing systems are known in the art. Today, storage virtualization and management is a separate entity than server virtualization and management. Different clients logical partitions (LPARs) associated with different servers access the same SAN storage. A client's LPARs on one server may not know if the SAN disk that it is trying to access is being used by some another client's LPAR belonging to some other server. This can cause data integrity issues and may potentially cause data corruption and client partition crashes.
Manually reinstalling and reconfiguring operating systems and applications at a secondary terminal is immensely time consuming and arduous task for a client (or a systems administrator) and also requires an amount of foresight that is not always possible (especially when terminals become unavailable due to hardware failure).
As cloud computing becomes more and more ubiquitous in the computer world, methods for providing enhanced functionality and greater up-time are required to continue to adequately serve commercial needs. Particularly for enterprise use, up-time is crucial. Therefore providing a client with access to their stored data with as close to 100% uptime is of the utmost importance.